run-self-service
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest user-supplied text to populate properties for self-service actions, such as provisioning cloud infrastructure or creating services.
- Ingestion points: User responses to prompts for action inputs defined in the execution process (Step 3,
SKILL.md). - Boundary markers: Instructions specify a structured form-like interaction and require a final confirmation from the user before execution (Step 4 and 5,
SKILL.md). - Capability inventory: The skill triggers external workflows and infrastructure changes via the
run_actiontool and manages Port entities through API calls inhelpers.py(upsert_entity). - Sanitization: The skill includes specific steps for input validation against schemas, including format checks for emails and URLs, to mitigate malformed data injection.
Audit Metadata